M AY 2 0 14
c o r p o r a t e f i n a n c e p r a c t i c e
Making better decisions about
the risks of capital projects
A handful of pragmatic tools can help managers decide which projects best fit
their portfolio and risk tolerance.
Martin Pergler and Never is the fear factor higher for managers than
Anders Rasmussen when they are making strategic investment
decisions on multibillion-dollar capital projects.
With such high stakes, we’ve seen many man-
agers prepare elaborate financial models to justify
potential projects. But when it comes down
to the final decision, especially when hard choices
need to be made among multiple opportunities,
they resort to less rigorous means—arbitrarily dis-
counting estimates of expected returns, for
example, or applying overly broad risk premiums.
There are more transparent ways to bring
assessments of risk into investment decisions. In
particular, we’ve found that some analytical
tools commonly employed by oil and gas compa-
nies can be particularly useful for players in other
capital-intensive industries, such as those
investing in projects with long lead times or those
investing in shorter-term projects that depend
on the economic cycle. The result can be a more
informed, data-driven discussion on a range
of possible outcomes. Of course, even these tools
are subject to assumptions that can be specu-
lative. But the insights they provide still produce
a more structured approach to making decisions
and a better dialogue about the trade-offs.
Some of the tools that follow may be familiar
to academics and even some practitioners. Many
companies use a subset of them in an ad hoc
fashion for particularly tricky decisions. The real

power comes from using them systematically,
however, leading to better decisions from a more
informed starting point: a fact-based depiction
of how much a company’s current performance is
at risk; a consistent assessment of each project’s
risks and returns; how those projects compare; and
how current and potential projects can be best
combined into a single portfolio.
Assess how much current performance
is at risk
Companies evaluating a new investment project
sometimes rush headlong into an assessment of
risks and returns of the project alone without
fully understanding the sources and magnitude of
the risks they already face. This isn’t surprising,
MoF 50since
perhaps, 2014managers naturally feel they know
Risk tools
their own for oilHowever,
business. and gasit does undermine
Exhibit 1 of 4
their ability to understand the potential results of a
new investment. Even a first-class evaluation
Exhibit 1 Overoptimistic assumptions of performance masked
a high likelihood of falling short of cash.
Projected cash flows vs cash needs, $ billion
0
1 2 3 4
2
of a new project only goes so far if managers
can’t compare it with the status quo or gauge the
incremental risk impact.
Consider, for example, what happened when
managers at one North American oil company
were evaluating a new investment. When
they quantified the company’s existing risk from
commodity-price uncertainty, transport
congestion, and regulatory developments, they
realized their assumptions were overly optimistic
about future cash flow in the new investment.
In fact, in a range of possible outcomes, there was
only a 5 percent chance that performance would
meet their base-case projections. Moreover, once
they mapped likely cash flows against capital
requirements and dividends, managers were
alarmed that they had only a 5 percent chance of
meeting their capital needs before the project’s
fourth year and on average wouldn’t meet them
Base case 5% case to Dividends and capital
Mean case 95% case commitments
5 6 7 8 9 10
Year
3

MoF 50 2014
Risk tools for oil and gas
Exhibit 2 of 4

Exhibit 2 Each project should be characterized in a simple

overview of risk-based economics.
Net-present-value (NPV) distribution, $ million Key metrics

5th Breakeven Expected Baseline Risk-corrected

percentile point NPV Baseline metrics metrics
Capital expenditure, 1,974 2,221
$ million

NPV, $ million 428 284

Internal rate of return, % 18.5 16.8

Payback period, years 10 10.9

Return on NPV/I,1 % 33 22
capital
RAROC,2 % 16

Key decomposition

Key risks NPV at risk, $ million

Capital-expenditure
–48 0 284 428 347
overrun
NPV at risk:
Crude price 310
$476 million
Probability to Probability to meet
break even = 92% baseline = 25% Execution delay 35

1 Ratio of NPV to investment.

2 Risk-adjusted return on capital = (expected NPV)/(planned investment + NPV at risk).

until year six (Exhibit 1). Fortunately, managers
were able to put the insight to good use;
they modified their strategic plan to make it
more resilient to risk.
Evaluate each project consistently
Once managers have a clear understanding of the
risks of their current portfolio of businesses,
they can drill down on risks in their proposed
projects and eliminate the need—and the
temptation—to adjust net present value (NPV) or
risk premiums arbitrarily. What’s needed is a
more consistent approach to evaluating project
economics and their risks, putting all potential
projects on equal footing. The cornerstone for such
an approach at one Middle Eastern oil company,
for example, is a standardized template for project
evaluation that features three main components:
a project’s risk-return profile at a glance, shown as
a probability distribution of project value; an
overview of standardized summary metrics for
risk and return; and an explicit description
of the sources of risk (Exhibit 2). The project team
specifies the basic economic drivers of the

project, but the central strategic-planning and risk
departments prescribe consistent key assumptions,
help to assess and challenge the risks identified,
and generally ensure that the method underlying
the analysis is robust.
This approach is pragmatic rather than mechanical.
A corporate-finance purist might challenge
the idea of a probability distribution of discounted
cash flows and the extent to which a chosen
discount rate accounts for the risk already, but in
practice, simplicity and transparency win over.
The project displayed in Exhibit 2 is one where the
economics are clearly worse than in the original
baseline proposal; indeed, this project is only 25
percent likely to meet that baseline. Never-
theless, it has more than a 90 percent chance of
breaking even. And even after taking into
account the potential need for additional invest-
ment after risks materialize, the project has
attractive returns.
Making this extra information about the distribu-
tion of outcomes available shifts the dialogue from
the typical go-no-go decision to a deeper
4
discussion about how to mitigate risk. In this
case, it is clearly worth exploring, for example, how
to reduce the likelihood of overruns in capital
expenditures in order to shift the entire probability
distribution to the right. This is likely considerably
easier to achieve if started early.
Prioritize projects by risk-adjusted returns
The reality in many industries, such as oil and gas,
is that companies have a large number of medium-
size projects, many of which are attractive on
a stand-alone basis—but they have limited capital
headroom to pursue them. It isn’t enough to
evaluate each project independently; they must
evaluate each relative to the others, too.
It’s not uncommon for managers to rank projects
based on some estimate of profitability or ratio
of NPV to investment. But since the challenge is to
figure out which projects are most likely to meet
expectations and which might require much
more scarce capital than initially anticipated, a
better approach is to evaluate them based on
a risk-adjusted ratio instead. At one multinational
downstream-focused oil company, managers
5

put this approach into practice by segmenting met an elevated hurdle rate were fast-tracked
projects based on an assessment of risk-adjusted without waiting for the annual prioritization
returns and then investing in new projects process. Projects in the middle, which would meet
up to the limit imposed by the amount of capital their cost of capital but did not exceed the
available (Exhibit 3). The first couple of itera- elevated hurdle rate, were rank ordered by their
tions of this process generated howls of protest. risk-adjusted returns. For these projects, ad
Project proponents repeatedly argued that hoc discussion can shift the rank ordering slightly.
their pet projects were strategic priorities—and But, more important, the exercise also quickly
that they urgently needed to go ahead without focuses attention on the handful of projects that
waiting until the system fixed all the other require nuanced consideration. That allows
projects’ numbers. managers to decide which ones they can move
forward safely, taking into account their risk and
The arguments receded as managers implemented the constraints of available capital.
MoF 50 2014
the process. Projects that clearly failed to meet
Risk tools
their cost for oil andlowest
of capital—the gas cutoff for risk- Determine the best overall portfolio
Exhibit 3 of 4
adjusted returns—were speedily declined or sent The approach above works well for companies that
back to the drawing board. Those that clearly seek to choose their investments from a large

Exhibit 3 Rank order projects by risk-adjusted returns to identify

which projects to fast-track and which to decline.
Projects rank ordered, Ongoing projects Fast-track approval Selectively approve Decline
by risk-adjusted returns

Soft- Hard-
capital capital
limit limit
Fast-track
hurdle

Cost of
capital1

Project A B C D E F G H I J K L M N O P
Available funds for deployment: External Cumulative capital
internal sources sources deployed

1 Weighted average cost of capital typically adjusted to eliminate double counting.

 6

number of similar medium-size projects. But they
may face opportunities quite different from their
existing portfolio—or they must weigh and set
project priorities for multiple strategies in different
directions—sometimes even before they’ve
identified specific projects. Usually this boils down
to a choice between doubling down on the kinds
of projects the company is already good at, even if
doing so increases exposure to concentrated
risk, and diversifying into an adjacent business.
Managers at another Middle Eastern energy
company, for example, had to decide whether to
invest in more upstream projects, where they
MoF 50 2014focused, or further develop the
were currently
Risk tools
company’s for oildownstream
nascent and gas portfolio.
Exhibit 4 of
Plotting the 4 in a risk-return graph allowed
options
managers to visualize the trade-offs of different
combinations of upstream and downstream
portfolio moves (Exhibit 4).
In this case, doubling down on upstream invest-
ments would have generated a portfolio with too
much risk. Emphasizing the downstream, with
only a slight increase in upstream activity, improved
on the status quo, but it left too much value
on the table. Through the exercise, managers dis-
covered that a moderate amount of commodity
hedging to manage margin volatility would allow
the company to combine a reduced upstream stake
with a significant downstream one. The risks of
the upstream and downstream investments were
quite diversified—and thus the company was able
to pursue a combined option with a more attractive
risk-return profile than either option alone.

Exhibit 4 Which set of strategic moves offers the best returns with
the right balance of upstream and downstream risk?
Return, 5-year operating cash flow, $ billion
Risk appetite
100
95 Current
90 + 75% downstream
+ 50% upstream with hedging
85
80 Current Current + maximum
+ 80% downstream upstream stake
75
+ 30% upstream
70
65
60
55
50
Current portfolio
45
40
0 5 10 15 20 25 30

Risk, 5-year cash flow at risk 95%, $ billion

7

The power of the approach lies in nudging Managing risk (and return) in capital-project
a strategic portfolio decision in a better direction and portfolio decisions will always be a challenge.
rather than analytically outsourcing port- But with an expanded set of tools, it is possible
folio construction. Managers in this case made no to focus risk-return decisions and enrich decision
attempt, for example, to calculate efficient fron- making, launching a dialogue about how to
tiers over a complex universe of portfolio choices. proactively manage those risks that matter most in
Oil companies that have tried such efforts have a more timely fashion.
often given up in frustration when their algorithms
suggested portfolio moves that were theoret-
ically precise but unachievable in practice, given
the likely response of counterparties and
other stakeholders.

Martin Pergler is a senior expert in McKinsey’s Montréal office, and Anders Rasmussen is a principal in the
London office. Copyright © 2014 McKinsey & Company. All rights reserved.